emory-libraries / dlp-lux

Discovery for the DLP Cor Repository

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

CircleCI Test Coverage

README

Discovery application for Emory's Cor repository.

  • Ruby version 2.7.5, Rails version 5.1

  • Blacklight 7

  • MySQL 5.7.22

Running locally

  1. Install and start up MySQL 5.7
  2. Clone the git repo: git clone git@github.com:emory-libraries/dlp-lux.git
  3. cd ./dlp-lux
  4. Install the required gems: bundle install
  5. Configure the MySQL development and test databases. This is accomplished by setting the environment variables that database.yml expects. In the root directory of the application, create a .env file and add the line DATABASE_USERNAME=root. (By default, MySQL has the username root with no password.) Also create a .env.development file and a .env.test file and add the lines DATABASE_NAME=dlp-lux_development and DATABASE_NAME=dlp-lux_test, respectively.
  6. In order to be able to sign into the application locally, the environment variable DATABASE_AUTH=true must be set in your development environment.
  • You must create a user via the rails console:
  bundle exec rails c
  u = User.new
  u.uid = "user"
  u.display_name = "User Name"
  u.email = "email@testdomain.com"
  u.password = "password"
  u.password_confirmation = "password"
  u.save
  1. In order to see objects with a visibility of "Rose High View", your IP must match an address on the server in config/reading_room_ips.yml. See the existing file for the example setup.
  2. Create and migrate the development database: rails db:create db:migrate
  3. Create and migrate the test database: RAILS_ENV=test rails db:create db:migrate
  4. Migrate the database: rails db:migrate
  5. Install UV and any other JS dependencies: yarn install
  6. Launch development instance of solr: solr_wrapper
  7. Launch test instance of solr: solr_wrapper --config config/solr_wrapper_test.yml
  8. Run the tests to ensure everything is working as expected: rspec spec
  9. Launch a rails server: rails server
  10. You should now be able to go to http://localhost:3000 and see the application

Running in production

  1. The IIIF_MANIFEST_URL environment variable needs to be set. This URL is the base URL for the Hyrax instance that serves the Work's IIIF manifest. An example: https://curate-qa.curationexperts.com/concern/curate_generic_works/.
  2. The THUMBNAIL_URL environment variable needs to be set. This URL is the base URL for the Hyrax instance that serves the Work's thumbnail images. An example: https://curate-qa.curationexperts.com

HTTP Password protection in production mode

In order to prevent search engine crawling of the system before it's ready to launch, we use HTTP password protection. This is set via environment variables. Set HTTP_PASSWORD_PROTECT='true' to enable this feature. Set HTTP_PASSWORD_PROTECT='false' to disable this feature. Set the login and password via environment variables HTTP_USERNAME and HTTP_PASSWORD

Deploying to Emory servers

  1. To deploy, first connect to the Emory VPN.
  2. run tki and receive your 12-hour AWS credentials.
  3. run the following command:
AWS_PROFILE=tki-aws-account-70-rhedcloud/RHEDcloudAdministratorRole BRANCH=<tag> bundle exec cap <env> deploy

where <tag> is either a version, like v0.4.32, or a branch, like master and where <env> is one of production, test, or arch.

Running test suite

  • Run rspec spec to run the default test suite
  • To run only specific tests, include
    • the directory, e.g. rspec spec/system,
    • or the path to the test, e.g. rspec spec/system/search_catalog_spec.rb,
    • or, if you want to be super specific, you can even add the line number the spec starts on at the beginning, e.g. rspec spec/system/search_catalog_spec.rb:178
    • Can also run with tags, e.g. rspec spec --tag relevancy:true

Run relevancy tests

  • Run PROD_LIKE_ENV=qa|test|arch|prod rake lux:relevancy
  • This rake task takes a prod-like environment as a variable and runs specs to ensure that search and indexing of active data are running as expected.

Run blacklight performance tests

  • Before your announced testing time window:
    • Install Apache Jmeter https://jmeter.apache.org/
    • From the project jmeter directory, run jmeter in GUI mode (jmeter with no command-line arguments on unix-like systems).
    • Open jmeter/blacklight.jmx from the file menu.
    • In the User Defined Variables panel, set the values of threads and loops to 1 for initial smoke-test run.
    • Check that the server name in the HTTP Request Defaults matches the server you want to run against (digital.library.emory.edu for production)
    • Add a username and password to the HTTP Authorization Manager if running against a system with HTTP basic auth in place.
    • "Save test plan as" and a temporary filename (blacklight-tmp.jmx for instance)
    • Run jmeter from the command line, replacing "trialname" with an unused filename: jmeter -n -t blacklight-tmp.jmx -l trialname.jtl -e -o report-trialname
    • Look for "Err: 0 (0.00%)" in the jmeter output; if there were errors, troubleshoot those before continuing to load testing.
  • During your announced testing time window:
    • Re-open your temporary copy of the test plan in the jmeter GUI
    • In the User Defined Variables panel, set the value of threads to the initial number of simultaneous users you want to simulate, and the value of loops to the number of times you want each simulated user to run through the suite of pages. Save your changes.
    • Run jmeter from the command line, replacing "trialname" with a new unused filename: jmeter -n -t blacklight-tmp.jmx -l trialname.jtl -e -o report-trialname
    • Repeat the previous two steps for each level of load to be simulated.

Run jmeter smoke test

  • Install and run Apache Jmeter in GUI mode as above.
  • Open jmeter/smoke-test.jmx from the file menu.
  • Add a username and password to the HTTP Authorization Manager if running against a system with HTTP basic auth in place.
  • Hit the green forward arrow to start the tests running.
  • View the results in "View Results Tree" (green with a checkmark for successes, red with an x for failures)

About

Discovery for the DLP Cor Repository


Languages

Language:Ruby 69.1%Language:HTML 17.5%Language:SCSS 8.7%Language:XSLT 4.0%Language:JavaScript 0.6%Language:CSS 0.2%